2003-07-21 07:48:48 +04:00
/*
2009-09-09 20:18:17 +04:00
* Copyright ( C ) 2003 - 2009 Kay Sievers < kay . sievers @ vrfy . org >
2003-07-21 07:48:48 +04:00
*
2008-09-10 04:40:42 +04:00
* This program is free software : you can redistribute it and / or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation , either version 2 of the License , or
* ( at your option ) any later version .
2003-07-21 07:48:48 +04:00
*
2008-09-10 04:40:42 +04:00
* This program is distributed in the hope that it will be useful ,
* but WITHOUT ANY WARRANTY ; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE . See the
* GNU General Public License for more details .
*
* You should have received a copy of the GNU General Public License
* along with this program . If not , see < http : //www.gnu.org/licenses/>.
2003-07-21 07:48:48 +04:00
*/
# include <stdlib.h>
# include <string.h>
# include <stdio.h>
2006-01-09 23:18:00 +03:00
# include <stddef.h>
2009-09-09 20:18:17 +04:00
# include <stdbool.h>
2003-07-21 07:48:48 +04:00
# include <fcntl.h>
# include <unistd.h>
# include <errno.h>
2006-02-03 15:52:37 +03:00
# include <grp.h>
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
# include <dirent.h>
2009-12-18 23:25:13 +03:00
# include <sys/time.h>
2003-10-23 06:39:54 +04:00
# include <sys/stat.h>
2003-11-24 07:56:18 +03:00
# include <sys/types.h>
2003-07-21 07:48:48 +04:00
# include "udev.h"
2004-10-06 10:39:05 +04:00
2007-06-21 04:28:09 +04:00
# define TMP_FILE_EXT ".udev-tmp"
2003-07-21 07:48:48 +04:00
2008-10-16 19:16:58 +04:00
int udev_node_mknod ( struct udev_device * dev , const char * file , dev_t devnum , mode_t mode , uid_t uid , gid_t gid )
{
struct udev * udev = udev_device_get_udev ( dev ) ;
2004-04-03 10:16:38 +04:00
struct stat stats ;
2008-07-11 01:54:28 +04:00
int preserve = 0 ;
int err = 0 ;
2004-04-03 10:16:38 +04:00
2008-10-16 19:16:58 +04:00
if ( major ( devnum ) = = 0 )
devnum = udev_device_get_devnum ( dev ) ;
if ( strcmp ( udev_device_get_subsystem ( dev ) , " block " ) = = 0 )
2005-06-18 12:57:10 +04:00
mode | = S_IFBLK ;
2006-01-09 23:18:00 +03:00
else
2005-06-18 12:57:10 +04:00
mode | = S_IFCHR ;
2009-05-04 06:52:31 +04:00
if ( file = = NULL )
2008-10-16 19:16:58 +04:00
file = udev_device_get_devnode ( dev ) ;
2007-06-21 04:28:09 +04:00
if ( lstat ( file , & stats ) = = 0 ) {
2008-10-16 19:16:58 +04:00
if ( ( ( stats . st_mode & S_IFMT ) = = ( mode & S_IFMT ) ) & & ( stats . st_rdev = = devnum ) ) {
info ( udev , " preserve file '%s', because it has correct dev_t \n " , file ) ;
2008-07-11 01:54:28 +04:00
preserve = 1 ;
2008-10-16 19:16:58 +04:00
udev_selinux_lsetfilecon ( udev , file , mode ) ;
2009-12-18 23:25:13 +03:00
/* update time stamp when we re-use the node, like on media change events */
utimes ( file , NULL ) ;
2008-07-11 01:54:28 +04:00
} else {
2009-04-19 05:25:04 +04:00
char file_tmp [ UTIL_PATH_SIZE + sizeof ( TMP_FILE_EXT ) ] ;
2008-10-16 19:16:58 +04:00
info ( udev , " atomically replace existing file '%s' \n " , file ) ;
2009-05-20 19:57:52 +04:00
util_strscpyl ( file_tmp , sizeof ( file_tmp ) , file , TMP_FILE_EXT , NULL ) ;
2008-07-11 01:54:28 +04:00
unlink ( file_tmp ) ;
2008-10-16 19:16:58 +04:00
udev_selinux_setfscreatecon ( udev , file_tmp , mode ) ;
err = mknod ( file_tmp , mode , devnum ) ;
udev_selinux_resetfscreatecon ( udev ) ;
2008-07-11 01:54:28 +04:00
if ( err ! = 0 ) {
2008-10-16 19:16:58 +04:00
err ( udev , " mknod(%s, %#o, %u, %u) failed: %m \n " ,
file_tmp , mode , major ( devnum ) , minor ( devnum ) ) ;
2008-07-11 01:54:28 +04:00
goto exit ;
}
err = rename ( file_tmp , file ) ;
if ( err ! = 0 ) {
2008-10-16 19:16:58 +04:00
err ( udev , " rename(%s, %s) failed: %m \n " , file_tmp , file ) ;
2008-07-11 01:54:28 +04:00
unlink ( file_tmp ) ;
}
2007-06-21 04:28:09 +04:00
}
2007-06-21 11:00:35 +04:00
} else {
2008-10-16 19:16:58 +04:00
info ( udev , " mknod(%s, %#o, (%u,%u)) \n " , file , mode , major ( devnum ) , minor ( devnum ) ) ;
2009-09-10 04:08:05 +04:00
do {
2009-10-28 13:37:20 +03:00
err = util_create_path ( udev , file ) ;
if ( err ! = 0 & & err ! = - ENOENT )
break ;
2009-09-10 04:08:05 +04:00
udev_selinux_setfscreatecon ( udev , file , mode ) ;
err = mknod ( file , mode , devnum ) ;
if ( err ! = 0 )
2009-10-28 13:37:20 +03:00
err = - errno ;
2009-09-10 04:08:05 +04:00
udev_selinux_resetfscreatecon ( udev ) ;
2009-10-28 13:37:20 +03:00
} while ( err = = - ENOENT ) ;
2008-07-11 01:54:28 +04:00
if ( err ! = 0 ) {
2008-10-16 19:16:58 +04:00
err ( udev , " mknod(%s, %#o, (%u,%u) failed: %m \n " , file , mode , major ( devnum ) , minor ( devnum ) ) ;
2008-07-11 01:54:28 +04:00
goto exit ;
}
2004-02-17 08:44:28 +03:00
}
2008-07-11 01:54:28 +04:00
if ( ! preserve | | stats . st_mode ! = mode ) {
2008-10-16 19:16:58 +04:00
info ( udev , " chmod(%s, %#o) \n " , file , mode ) ;
2008-07-11 01:54:28 +04:00
err = chmod ( file , mode ) ;
if ( err ! = 0 ) {
2008-10-16 19:16:58 +04:00
err ( udev , " chmod(%s, %#o) failed: %m \n " , file , mode ) ;
2008-07-11 01:54:28 +04:00
goto exit ;
}
2004-02-17 08:44:28 +03:00
}
2008-07-11 01:54:28 +04:00
if ( ! preserve | | stats . st_uid ! = uid | | stats . st_gid ! = gid ) {
2008-10-16 19:16:58 +04:00
info ( udev , " chown(%s, %u, %u) \n " , file , uid , gid ) ;
2008-07-11 01:54:28 +04:00
err = chown ( file , uid , gid ) ;
if ( err ! = 0 ) {
2008-10-16 19:16:58 +04:00
err ( udev , " chown(%s, %u, %u) failed: %m \n " , file , uid , gid ) ;
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
goto exit ;
2004-02-17 08:44:28 +03:00
}
}
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
exit :
2008-07-11 01:54:28 +04:00
return err ;
2004-02-17 08:44:28 +03:00
}
2008-10-16 19:16:58 +04:00
static int node_symlink ( struct udev * udev , const char * node , const char * slink )
2006-06-14 18:32:52 +04:00
{
2007-06-21 04:28:09 +04:00
struct stat stats ;
2008-10-26 04:31:54 +03:00
char target [ UTIL_PATH_SIZE ] ;
2009-05-20 19:57:52 +04:00
char * s ;
size_t l ;
2008-09-10 23:50:21 +04:00
char slink_tmp [ UTIL_PATH_SIZE + sizeof ( TMP_FILE_EXT ) ] ;
2007-03-15 05:54:12 +03:00
int i = 0 ;
int tail = 0 ;
2008-10-16 19:16:58 +04:00
int err = 0 ;
2006-06-14 18:32:52 +04:00
2007-03-15 05:54:12 +03:00
/* use relative link */
2008-10-26 04:31:54 +03:00
target [ 0 ] = ' \0 ' ;
2007-03-15 05:54:12 +03:00
while ( node [ i ] & & ( node [ i ] = = slink [ i ] ) ) {
if ( node [ i ] = = ' / ' )
tail = i + 1 ;
i + + ;
}
2009-05-20 19:57:52 +04:00
s = target ;
l = sizeof ( target ) ;
2007-03-15 05:54:12 +03:00
while ( slink [ i ] ! = ' \0 ' ) {
if ( slink [ i ] = = ' / ' )
2009-05-20 19:57:52 +04:00
l = util_strpcpy ( & s , l , " ../ " ) ;
2007-03-15 05:54:12 +03:00
i + + ;
}
2009-05-20 19:57:52 +04:00
l = util_strscpy ( s , l , & node [ tail ] ) ;
if ( l = = 0 ) {
err = - EINVAL ;
goto exit ;
}
2007-03-15 05:54:12 +03:00
2007-06-21 04:28:09 +04:00
/* preserve link with correct target, do not replace node of other device */
if ( lstat ( slink , & stats ) = = 0 ) {
if ( S_ISBLK ( stats . st_mode ) | | S_ISCHR ( stats . st_mode ) ) {
struct stat stats2 ;
2008-10-16 19:16:58 +04:00
info ( udev , " found existing node instead of symlink '%s' \n " , slink ) ;
2007-06-21 04:28:09 +04:00
if ( lstat ( node , & stats2 ) = = 0 ) {
if ( ( stats . st_mode & S_IFMT ) = = ( stats2 . st_mode & S_IFMT ) & &
2009-09-09 20:18:17 +04:00
stats . st_rdev = = stats2 . st_rdev & & stats . st_ino ! = stats2 . st_ino ) {
2008-10-16 19:16:58 +04:00
info ( udev , " replace device node '%s' with symlink to our node '%s' \n " ,
slink , node ) ;
2007-06-21 04:28:09 +04:00
} else {
2008-10-16 19:16:58 +04:00
err ( udev , " device node '%s' already exists, "
" link to '%s' will not overwrite it \n " ,
slink , node ) ;
2007-06-21 04:28:09 +04:00
goto exit ;
}
}
} else if ( S_ISLNK ( stats . st_mode ) ) {
2008-09-10 23:50:21 +04:00
char buf [ UTIL_PATH_SIZE ] ;
2009-05-20 19:57:52 +04:00
int len ;
2007-06-21 04:28:09 +04:00
2008-11-01 22:05:41 +03:00
dbg ( udev , " found existing symlink '%s' \n " , slink ) ;
2007-06-21 04:28:09 +04:00
len = readlink ( slink , buf , sizeof ( buf ) ) ;
if ( len > 0 ) {
buf [ len ] = ' \0 ' ;
if ( strcmp ( target , buf ) = = 0 ) {
2008-10-16 19:16:58 +04:00
info ( udev , " preserve already existing symlink '%s' to '%s' \n " ,
slink , target ) ;
udev_selinux_lsetfilecon ( udev , slink , S_IFLNK ) ;
2009-12-18 23:25:13 +03:00
lutimes ( slink , NULL ) ;
2007-06-21 04:28:09 +04:00
goto exit ;
}
}
2006-06-14 18:32:52 +04:00
}
2007-06-21 11:00:35 +04:00
} else {
2008-10-16 19:16:58 +04:00
info ( udev , " creating symlink '%s' to '%s' \n " , slink , target ) ;
2009-09-10 04:08:05 +04:00
do {
2009-10-28 13:37:20 +03:00
err = util_create_path ( udev , slink ) ;
if ( err ! = 0 & & err ! = - ENOENT )
break ;
2009-09-10 04:08:05 +04:00
udev_selinux_setfscreatecon ( udev , slink , S_IFLNK ) ;
err = symlink ( target , slink ) ;
if ( err ! = 0 )
2009-10-28 13:37:20 +03:00
err = - errno ;
2009-09-10 04:08:05 +04:00
udev_selinux_resetfscreatecon ( udev ) ;
2009-10-28 13:37:20 +03:00
} while ( err = = - ENOENT ) ;
2008-10-16 19:16:58 +04:00
if ( err = = 0 )
2007-06-21 11:00:35 +04:00
goto exit ;
2006-06-14 18:32:52 +04:00
}
2008-10-16 19:16:58 +04:00
info ( udev , " atomically replace '%s' \n " , slink ) ;
2009-05-20 19:57:52 +04:00
util_strscpyl ( slink_tmp , sizeof ( slink_tmp ) , slink , TMP_FILE_EXT , NULL ) ;
2007-06-25 18:03:11 +04:00
unlink ( slink_tmp ) ;
2009-09-10 04:08:05 +04:00
do {
2009-10-28 13:37:20 +03:00
err = util_create_path ( udev , slink_tmp ) ;
if ( err ! = 0 & & err ! = - ENOENT )
break ;
udev_selinux_setfscreatecon ( udev , slink_tmp , S_IFLNK ) ;
2009-09-10 04:08:05 +04:00
err = symlink ( target , slink_tmp ) ;
if ( err ! = 0 )
2009-10-28 13:37:20 +03:00
err = - errno ;
2009-09-10 04:08:05 +04:00
udev_selinux_resetfscreatecon ( udev ) ;
2009-10-28 13:37:20 +03:00
} while ( err = = - ENOENT ) ;
2008-10-16 19:16:58 +04:00
if ( err ! = 0 ) {
err ( udev , " symlink(%s, %s) failed: %m \n " , target , slink_tmp ) ;
2007-06-21 04:28:09 +04:00
goto exit ;
}
2008-10-16 19:16:58 +04:00
err = rename ( slink_tmp , slink ) ;
if ( err ! = 0 ) {
err ( udev , " rename(%s, %s) failed: %m \n " , slink_tmp , slink ) ;
2007-06-21 04:28:09 +04:00
unlink ( slink_tmp ) ;
}
2006-06-14 18:32:52 +04:00
exit :
2008-10-16 19:16:58 +04:00
return err ;
2006-06-14 18:32:52 +04:00
}
2009-09-09 20:18:17 +04:00
/* find device node of device with highest priority */
static const char * link_find_prioritized ( struct udev_device * dev , bool add , const char * stackdir , char * buf , size_t bufsize )
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
{
2009-09-09 20:18:17 +04:00
struct udev * udev = udev_device_get_udev ( dev ) ;
2008-10-16 19:16:58 +04:00
DIR * dir ;
2009-09-09 20:18:17 +04:00
int priority = 0 ;
const char * target = NULL ;
if ( add ) {
priority = udev_device_get_devlink_priority ( dev ) ;
util_strscpy ( buf , bufsize , udev_device_get_devnode ( dev ) ) ;
target = buf ;
2008-10-16 19:16:58 +04:00
}
2008-10-16 21:23:07 +04:00
2009-09-09 20:18:17 +04:00
dir = opendir ( stackdir ) ;
if ( dir = = NULL )
return target ;
for ( ; ; ) {
struct udev_device * dev_db ;
2009-08-19 22:49:49 +04:00
struct dirent * dent ;
char devpath [ UTIL_PATH_SIZE ] ;
char syspath [ UTIL_PATH_SIZE ] ;
2009-09-09 20:18:17 +04:00
ssize_t len ;
2008-10-16 19:16:58 +04:00
2009-08-19 22:49:49 +04:00
dent = readdir ( dir ) ;
if ( dent = = NULL | | dent - > d_name [ 0 ] = = ' \0 ' )
2008-10-16 19:16:58 +04:00
break ;
2009-08-19 22:49:49 +04:00
if ( dent - > d_name [ 0 ] = = ' . ' )
2008-10-16 19:16:58 +04:00
continue ;
2009-09-09 20:18:17 +04:00
dbg ( udev , " found '%s/%s' \n " , stackdir , dent - > d_name ) ;
2009-08-19 22:49:49 +04:00
len = readlinkat ( dirfd ( dir ) , dent - > d_name , devpath , sizeof ( devpath ) ) ;
2009-09-09 20:18:17 +04:00
if ( len < = 0 | | len = = ( ssize_t ) sizeof ( devpath ) )
2009-08-19 22:49:49 +04:00
continue ;
devpath [ len ] = ' \0 ' ;
util_strscpyl ( syspath , sizeof ( syspath ) , udev_get_sys_path ( udev ) , devpath , NULL ) ;
2009-09-09 20:18:17 +04:00
info ( udev , " found '%s' claiming '%s' \n " , syspath , stackdir ) ;
/* did we find ourself? */
if ( strcmp ( udev_device_get_syspath ( dev ) , syspath ) = = 0 )
continue ;
dev_db = udev_device_new_from_syspath ( udev , syspath ) ;
if ( dev_db ! = NULL ) {
const char * devnode ;
devnode = udev_device_get_devnode ( dev_db ) ;
if ( devnode ! = NULL ) {
dbg ( udev , " compare priority of '%s'(%i) > '%s'(%i) \n " , target , priority ,
udev_device_get_devnode ( dev_db ) , udev_device_get_devlink_priority ( dev_db ) ) ;
if ( target = = NULL | | udev_device_get_devlink_priority ( dev_db ) > priority ) {
info ( udev , " '%s' claims priority %i for '%s' \n " ,
syspath , udev_device_get_devlink_priority ( dev_db ) , stackdir ) ;
priority = udev_device_get_devlink_priority ( dev_db ) ;
util_strscpy ( buf , bufsize , devnode ) ;
target = buf ;
}
}
udev_device_unref ( dev_db ) ;
}
2008-10-16 19:16:58 +04:00
}
2009-11-03 02:03:43 +03:00
closedir ( dir ) ;
2009-09-09 20:18:17 +04:00
return target ;
2008-10-16 19:16:58 +04:00
}
2009-09-09 20:18:17 +04:00
/* manage "stack of names" with possibly specified device priorities */
static void link_update ( struct udev_device * dev , const char * slink , bool add )
2008-10-16 19:16:58 +04:00
{
struct udev * udev = udev_device_get_udev ( dev ) ;
2009-09-09 20:18:17 +04:00
char name_enc [ UTIL_PATH_SIZE ] ;
char filename [ UTIL_PATH_SIZE * 2 ] ;
char dirname [ UTIL_PATH_SIZE ] ;
const char * target ;
char buf [ UTIL_PATH_SIZE ] ;
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
2008-11-01 22:05:41 +03:00
dbg ( udev , " update symlink '%s' of '%s' \n " , slink , udev_device_get_syspath ( dev ) ) ;
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
2009-09-09 20:18:17 +04:00
util_path_encode ( & slink [ strlen ( udev_get_dev_path ( udev ) ) + 1 ] , name_enc , sizeof ( name_enc ) ) ;
snprintf ( dirname , sizeof ( dirname ) , " %s/.udev/links/%s " , udev_get_dev_path ( udev ) , name_enc ) ;
snprintf ( filename , sizeof ( filename ) , " %s/%c%u:%u " , dirname ,
strcmp ( udev_device_get_subsystem ( dev ) , " block " ) = = 0 ? ' b ' : ' c ' ,
major ( udev_device_get_devnum ( dev ) ) , minor ( udev_device_get_devnum ( dev ) ) ) ;
if ( ! add ) {
dbg ( udev , " removing index: '%s' \n " , filename ) ;
unlink ( filename ) ;
util_delete_path ( udev , filename ) ;
}
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
2009-09-09 20:18:17 +04:00
target = link_find_prioritized ( dev , add , dirname , buf , sizeof ( buf ) ) ;
if ( target = = NULL ) {
2008-10-16 19:16:58 +04:00
info ( udev , " no reference left, remove '%s' \n " , slink ) ;
2009-01-15 22:20:16 +03:00
unlink ( slink ) ;
util_delete_path ( udev , slink ) ;
2009-09-09 20:18:17 +04:00
} else {
info ( udev , " creating link '%s' to '%s' \n " , slink , target ) ;
node_symlink ( udev , target , slink ) ;
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
}
2009-09-09 20:18:17 +04:00
if ( add ) {
2009-10-28 13:37:20 +03:00
int err ;
2009-09-09 20:18:17 +04:00
dbg ( udev , " creating index: '%s' \n " , filename ) ;
2009-09-10 04:08:05 +04:00
do {
2009-10-28 13:37:20 +03:00
err = util_create_path ( udev , filename ) ;
if ( err ! = 0 & & err ! = - ENOENT )
break ;
err = symlink ( udev_device_get_devpath ( dev ) , filename ) ;
if ( err ! = 0 )
err = - errno ;
} while ( err = = - ENOENT ) ;
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
}
}
2009-01-15 22:20:16 +03:00
void udev_node_update_old_links ( struct udev_device * dev , struct udev_device * dev_old )
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
{
2008-10-16 19:16:58 +04:00
struct udev * udev = udev_device_get_udev ( dev ) ;
struct udev_list_entry * list_entry ;
/* update possible left-over symlinks */
udev_list_entry_foreach ( list_entry , udev_device_get_devlinks_list_entry ( dev_old ) ) {
2008-11-05 06:45:23 +03:00
const char * name = udev_list_entry_get_name ( list_entry ) ;
2008-10-16 19:16:58 +04:00
struct udev_list_entry * list_entry_current ;
2008-11-01 22:05:41 +03:00
int found ;
2008-10-16 19:16:58 +04:00
2008-11-24 04:21:12 +03:00
/* check if old link name still belongs to this device */
2008-11-01 22:05:41 +03:00
found = 0 ;
2008-10-16 19:16:58 +04:00
udev_list_entry_foreach ( list_entry_current , udev_device_get_devlinks_list_entry ( dev ) ) {
2008-11-05 06:45:23 +03:00
const char * name_current = udev_list_entry_get_name ( list_entry_current ) ;
2008-11-24 04:21:12 +03:00
if ( strcmp ( name , name_current ) = = 0 ) {
2008-11-01 22:05:41 +03:00
found = 1 ;
break ;
}
2008-10-16 19:16:58 +04:00
}
2008-11-01 22:05:41 +03:00
if ( found )
continue ;
2008-11-24 04:21:12 +03:00
2009-01-16 23:14:13 +03:00
info ( udev , " update old name, '%s' no longer belonging to '%s' \n " ,
name , udev_device_get_devpath ( dev ) ) ;
2009-09-09 20:18:17 +04:00
link_update ( dev , name , 0 ) ;
priority based symlink handling
Symlinks can have priorities now, the priority is assigned to the device
and specified with OPTIONS="link_priority=100". Devices with higher
priorities overwrite the symlinks of devices with lower priorities.
If the device, that currently owns the link goes away, the symlink
will be removed, and recreated, pointing to the next device with the
highest actual priority.
This should solve the issue, that inserting an USB-stick may overwrite the
/dev/disk/by-id/-link of another disk, and removes the entire link after the
USB-stick is disconnected. If no priorities are specified, the new link will
overwrite the current one, and if the device goes away, it will restore
the old link. It should be possible to assign lower priorities to removable
devices, if needed.
In multipath setups, we see several devices, which all connect to the same
volume, and therefore all try to create the same metadata-links. The
different path-devices are combined into one device-mapper device, which also
contains the same metadata. It should be possible, to assign multipath-table
device-mapper devices a higher priority, so path-devices that appear and
disappear, will not overwrite or delete the device-mapper device links.
2007-03-18 14:51:57 +03:00
}
}
2009-01-15 22:20:16 +03:00
int udev_node_add ( struct udev_device * dev , mode_t mode , uid_t uid , gid_t gid )
2004-03-02 10:31:06 +03:00
{
2008-10-16 19:16:58 +04:00
struct udev * udev = udev_device_get_udev ( dev ) ;
2005-03-05 07:35:31 +03:00
int i ;
2008-10-16 19:16:58 +04:00
int num ;
struct udev_list_entry * list_entry ;
int err = 0 ;
2006-04-06 00:29:33 +04:00
2008-10-16 19:16:58 +04:00
info ( udev , " creating device node '%s', devnum=%d:%d, mode=%#o, uid=%d, gid=%d \n " ,
udev_device_get_devnode ( dev ) ,
major ( udev_device_get_devnum ( dev ) ) , minor ( udev_device_get_devnum ( dev ) ) ,
mode , uid , gid ) ;
2006-01-25 04:18:13 +03:00
2009-01-15 22:20:16 +03:00
if ( udev_node_mknod ( dev , NULL , makedev ( 0 , 0 ) , mode , uid , gid ) ! = 0 ) {
err = - 1 ;
goto exit ;
}
2006-01-25 04:18:13 +03:00
[PATCH] netdev - udevdb+dev.d changes
Here is a patch to change the netdev handling in the database and for
the dev.d/ calls. I applies on top of the udevd.patch, cause klibc has
no sysinfo().
o netdev's are also put into our database now. I want this for the
udevruler gui to get a list of all handled devices.
All devices in the db are stamped with the system uptime value at
the creation time. 'udevinfo -d' prints it.
o the DEVPATH value is the key for udevdb, but if we rename
a netdev, the name is replaced in the kernel, so we add
the changed name to the db to match with the remove event.
NOTE: The dev.d/ scripts still get the original name from the
hotplug call. Should we replace DEVPATH with the new name too?
o We now only add a device to the db, if we have successfully created
the main node or successfully renamed a netdev. This is the main part
of the patch, cause I needed to clean the retval passing trough all
the functions used for node creation.
o DEVNODE sounds a bit ugly for netdev's so I exported DEVNAME too.
Can we change the name?
o I've added a UDEV_NO_DEVD to possibly skip the script execution
and used it in udev-test.pl.
udevstart is the same horror now, if you have scripts with logging
statements in dev.d/ it takes minutes to finish, can we skip the
scripts here too?
o The get_device_type() function is changed to be more strict, cause
'udevinfo -a -p /block/' gets a class device for it and tries to
print the major/minor values.
o bugfix, the RESULT value has now a working newline removal and a test
for this case.
2004-04-01 11:12:57 +04:00
/* create all_partitions if requested */
2008-10-16 19:16:58 +04:00
num = udev_device_get_num_fake_partitions ( dev ) ;
if ( num > 0 ) {
info ( udev , " creating device partition nodes '%s[1-%i]' \n " , udev_device_get_devnode ( dev ) , num ) ;
2009-01-15 22:20:16 +03:00
for ( i = 1 ; i < = num ; i + + ) {
char partitionname [ UTIL_PATH_SIZE ] ;
dev_t part_devnum ;
snprintf ( partitionname , sizeof ( partitionname ) , " %s%d " ,
udev_device_get_devnode ( dev ) , i ) ;
partitionname [ sizeof ( partitionname ) - 1 ] = ' \0 ' ;
part_devnum = makedev ( major ( udev_device_get_devnum ( dev ) ) ,
minor ( udev_device_get_devnum ( dev ) ) + i ) ;
udev_node_mknod ( dev , partitionname , part_devnum , mode , uid , gid ) ;
2004-02-17 08:44:28 +03:00
}
2003-12-07 20:12:07 +03:00
}
2008-10-16 19:16:58 +04:00
/* create/update symlinks, add symlinks to name index */
udev_list_entry_foreach ( list_entry , udev_device_get_devlinks_list_entry ( dev ) ) {
2009-09-10 03:35:48 +04:00
if ( udev_list_entry_get_flag ( list_entry ) )
2009-09-09 20:18:17 +04:00
/* simple unmanaged link name */
node_symlink ( udev , udev_device_get_devnode ( dev ) , udev_list_entry_get_name ( list_entry ) ) ;
2009-09-10 03:35:48 +04:00
else
2009-09-09 20:18:17 +04:00
link_update ( dev , udev_list_entry_get_name ( list_entry ) , 1 ) ;
2008-10-16 19:16:58 +04:00
}
2006-04-06 00:29:33 +04:00
exit :
2008-10-16 19:16:58 +04:00
return err ;
2003-07-21 07:48:48 +04:00
}
2009-04-26 15:59:39 +04:00
int udev_node_remove ( struct udev_device * dev )
2006-04-13 00:08:05 +04:00
{
2008-10-16 19:16:58 +04:00
struct udev * udev = udev_device_get_udev ( dev ) ;
struct udev_list_entry * list_entry ;
const char * devnode ;
2008-09-10 23:50:21 +04:00
char partitionname [ UTIL_PATH_SIZE ] ;
2006-04-13 00:08:05 +04:00
struct stat stats ;
2008-10-16 19:16:58 +04:00
int err = 0 ;
2006-04-13 00:08:05 +04:00
int num ;
2008-10-16 19:16:58 +04:00
/* remove,update symlinks, remove symlinks from name index */
2009-09-09 20:18:17 +04:00
udev_list_entry_foreach ( list_entry , udev_device_get_devlinks_list_entry ( dev ) )
link_update ( dev , udev_list_entry_get_name ( list_entry ) , 0 ) ;
2008-10-16 19:16:58 +04:00
devnode = udev_device_get_devnode ( dev ) ;
if ( devnode = = NULL )
return 0 ;
if ( stat ( devnode , & stats ) ! = 0 ) {
info ( udev , " device node '%s' not found \n " , devnode ) ;
2007-12-29 19:19:06 +03:00
return 0 ;
2006-04-06 00:29:33 +04:00
}
2008-10-16 19:16:58 +04:00
if ( stats . st_rdev ! = udev_device_get_devnum ( dev ) ) {
info ( udev , " device node '%s' points to a different device, skip removal \n " , devnode ) ;
2006-04-06 00:29:33 +04:00
return - 1 ;
}
2008-10-16 19:16:58 +04:00
info ( udev , " removing device node '%s' \n " , devnode ) ;
2009-01-15 22:20:16 +03:00
err = util_unlink_secure ( udev , devnode ) ;
2008-10-16 19:16:58 +04:00
if ( err )
return err ;
2006-04-06 00:29:33 +04:00
2008-10-16 19:16:58 +04:00
num = udev_device_get_num_fake_partitions ( dev ) ;
2006-04-06 00:29:33 +04:00
if ( num > 0 ) {
2006-04-13 00:08:05 +04:00
int i ;
2008-10-16 19:16:58 +04:00
info ( udev , " removing all_partitions '%s[1-%i]' \n " , devnode , num ) ;
2007-03-16 23:15:54 +03:00
if ( num > 255 )
2006-04-06 00:29:33 +04:00
return - 1 ;
for ( i = 1 ; i < = num ; i + + ) {
2008-10-16 19:16:58 +04:00
snprintf ( partitionname , sizeof ( partitionname ) , " %s%d " , devnode , i ) ;
2006-04-06 00:29:33 +04:00
partitionname [ sizeof ( partitionname ) - 1 ] = ' \0 ' ;
2009-01-15 22:20:16 +03:00
util_unlink_secure ( udev , partitionname ) ;
2004-10-19 06:28:39 +04:00
}
[PATCH] hmm, handle net devices with udev?
Hmm, Arndt Bergmann sent a patch like this one a few weeks ago and
I want to bring the question back, if we want to handle net device
naming with udev.
With this patch it is actually possible to specify something like this
in udev.rules:
KERNEL="dummy*", SYSFS{address}="00:00:00:00:00:00", SYSFS{features}="0x0", NAME="blind%n"
KERNEL="eth*", SYSFS{address}="00:0d:60:77:30:91", NAME="private"
and you will get:
[root@pim udev.kay]# cat /proc/net/dev
Inter-| Receive | Transmit
face |bytes packets errs drop fifo frame compressed multicast|bytes packets errs drop fifo colls carrier compressed
lo: 1500 30 0 0 0 0 0 0 1500 30 0 0 0 0 0 0
private: 278393 1114 0 0 0 0 0 0 153204 1468 0 0 0 0 0 0
sit0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
blind0: 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
The udevinfo program is also working:
[root@pim udev.kay]# ./udevinfo -a -p /sys/class/net/private
looking at class device '/sys/class/net/private':
SYSFS{addr_len}="6"
SYSFS{address}="00:0d:60:77:30:91"
SYSFS{broadcast}="ff:ff:ff:ff:ff:ff"
SYSFS{features}="0x3a9"
SYSFS{flags}="0x1003"
SYSFS{ifindex}="2"
SYSFS{iflink}="2"
SYSFS{mtu}="1500"
SYSFS{tx_queue_len}="1000"
SYSFS{type}="1"
follow the class device's "device"
looking at the device chain at '/sys/devices/pci0000:00/0000:00:1e.0/0000:02:01.0':
BUS="pci"
ID="0000:02:01.0"
SYSFS{class}="0x020000"
SYSFS{detach_state}="0"
SYSFS{device}="0x101e"
SYSFS{irq}="11"
SYSFS{subsystem_device}="0x0549"
SYSFS{subsystem_vendor}="0x1014"
SYSFS{vendor}="0x8086"
The matching device will be renamed to the given name. The device name
will not be put into the udev database, cause the kernel renames the
device and the sysfs name disappears.
I like it, cause it plugs in nicely. We have all the naming features
and sysfs queries and walks inside of udev. The sysfs timing races
are already solved and the management tools are working for net devices
too. nameif can only match the MAC address now. udev can match any sysfs
value of the device tree the net device is connected to.
But right, net devices do not have device nodes :)
2004-03-25 10:19:39 +03:00
}
2008-10-18 22:12:55 +04:00
util_delete_path ( udev , devnode ) ;
2008-10-16 19:16:58 +04:00
return err ;
2003-07-21 07:48:48 +04:00
}